THE UNIVERSITY OF BRITISH COLUMBIA
Background: A diagnosis of asthma is considered an independent risk factor for COPD. However, little is known about health services utilization patterns in COPD patients with a history of asthma in comparison to those without such history, especially with regard to comorbid conditions. Objectives: To estimate the excess costs of COPD with a history of asthma (COPD+asthma) versus COPD without such history (COPD-only); to estimate excess costs attributable to inpatient care, outpatient care, medications, and community care; and to estimate excess costs attributable to comorbid conditions. Methods: We used vital statistics, inpatient and outpatient encounters, filled prescription records, and community care data for British Columbia, Canada, from 1997 to 2012, to create a propensity-score-matched cohort of COPD+asthma and COPD-only patients. We calculated and compared the excess medical costs (in 2012 Canadian Dollars [$]) between the two groups based on billing information. Commodities were ascertained from the inpatient and outpatient records and were classified based on major categories of the international classification of diseases, 10th revision. Results: The final sample consisted of 22,565 individuals within each group (mean age at baseline 67.9, 57.0% female, average follow-up 4.07 years). Excess costs of COPD+asthma were $540.7/PY (95%CI $301.7‒$779.8, P<0.001). Costs of medications ($657.9/PY,P<0.001) and outpatient services ($127.6/PY,P<0.001) were higher in COPD+asthma but costs of hospitalizations were lower (-$271.0/PY,P=0.002) while community care costs were similar between the two groups (P=0.257). Excess costs of respiratory-related conditions were $856.2/PY (P<0.001), with $552.6/PY due to respiratory-related medications (P<0.001); costs of all other conditions combined were lower in COPD+asthma, mainly due to lower costs of cardiovascular diseases (CVD, -$201.8/PY, P<0.001). Conclusion: COPD patients with a previous history of asthma consume more health care resources than those with COPD alone, but there are important differences in cost components and costs attributable to comorbid conditions. Further research is required to examine whether the lower costs of CVD in these patients is due to lower levels of related risk factors or intrinsic differences in COPD phenotypes.